49
Table A3: |
Robustness Checks: |
Semi-log specification an |
different IV strategy | |||||
(1) |
(2) |
(3) |
(4) |
(5) |
(6) |
(7) |
(8) | |
Dep.var. |
ln u |
ln u |
u |
u |
u |
u |
ln u |
u |
OLS |
IV |
IV |
OLS |
OLS |
IV |
OLS |
OLS | |
Instruments / Proxy |
bt*-1, pmrt*-1 |
b*.pmr* |
* |
bt*-1, pmrt*-1 | ||||
ln u*, u* |
0.018 |
0.022c |
0.080a |
0.075a |
0.086a |
0.087a | ||
(0.013) |
(0.013) |
(0.028) |
(0.027) |
(0.030) |
(0.032) | |||
b* |
0.002a |
0.009a | ||||||
(0.001) |
(0.003) | |||||||
b |
0.017a |
0.018a |
0.085a |
0.099a |
0.078a |
0.095a |
0.018a |
0.079a |
(0.003) |
(0.003) |
(0.018) |
(0.020) |
(0.020) |
(0.019) |
(0.003) |
(0.020) | |
PMR |
0.184a |
0.198a |
0.875a |
0.939a |
0.843a |
0.920a |
0.182a |
0.821a |
(0.037) |
(0.033) |
(0.196) |
(0.215) |
(0.210) |
(0.203) |
(0.036) |
(0.206) | |
Union density |
-0.003 |
-0.003 |
-0.004 |
-0.003 |
-0.005 |
-0.004 |
-0.003 |
-0.005 |
(0.003) |
(0.003) |
(0.022) |
(0.023) |
(0.023) |
(0.022) |
(0.003) |
(0.023) | |
High corporatism |
-0.129b |
-0.150a |
-1.619a |
-1.865a |
-1.527a |
-1.781a |
-0.124b |
-1.525a |
(0.057) |
(0.056) |
(0.390) |
(0.427) |
(0.415) |
(0.390) |
(0.057) |
(0.415) | |
EPL |
-0.012 |
0.007 |
-0.413 |
-0.300 |
-0.458 |
-0.326 |
-0.020 |
-0.461 |
(0.054) |
(0.049) |
(0.342) |
(0.375) |
(0.366) |
(0.351) |
(0.056) |
(0.371) | |
EU dummy |
0.724c |
0.438 |
0.732c | |||||
(0.383) |
(0.355) |
(0.383) | ||||||
gap |
-0.091a |
-0.090a |
-0.621a |
-0.609a |
-0.639a |
-0.617a |
-0.091a |
-0.641a |
(0.009) |
(0.008) |
(0.041) |
(0.042) |
(0.043) |
(0.039) |
(0.009) |
(0.043) | |
* |
-0.008 |
-0.002 |
0.021 |
0.015 |
0.034 |
0.073 |
-0.008 |
-0.015 |
(0.015) |
(0.014) |
(0.070) |
(0.067) |
(0.073) |
(0.070) |
(0.015) |
(0.070) | |
2nd stage statistics | ||||||||
RMSE |
0.210 |
0.188 |
1.046 |
1.095 |
1.123 |
1.008 |
0.210 |
1.124 |
adj. R2 |
0.893 |
0.896 |
0.931 |
0.935 |
0.932 |
0.936 |
0.893 |
0.931 |
F |
72.41 |
158.7 |
150.0 |
73.06 |
149.7 | |||
1st stage statistics (p-valus) | ||||||||
partial R2 |
0.633 |
0.972 |
0.753 | |||||
χ2 -overidentification |
0.0190 |
0.680 |
0.970 | |||||
χ2 -endogeneity |
0.695 |
0.623 |
0.886 |
Robust standard errors in parentheses,ap < 0.01,bp < 0.05,cp < 0.1. Number of observations: 397 in OLS and 374
in IV regressions. All regressions contain a full set of country fixed-effects, year dummies, and an array of orthogonal
shocks (TFP, terms of trade, real interest rate, and labor demand shocks) as additional controls for business cycle
comovements. Trade-weighted averages for foreign variables (denoted by asterisks) are computed using α1 = α2 =
1,δ = 1. In IV regressions, the foreign unemployment rate u* is instrumented by b*-1, PMR*-1 and gap*-1. The
χ2-endogeneity test tests the null that u* is exogenous (and rejects in all presented IV models). Overidentification
is tested for using the Wooldridge robust Score test (invalidity of instruments or model misspecification is rejected
in all presented IV models).